In conventional systems, users have the ability to enter search queries for performing searches for a wide variety of information related to media. Because modern search engines can effectively associate a relevant result with one or two significant keywords, users may be able to quickly find relevant results for search queries describing simple concepts, which typically require a small number of significant keywords to describe. However, in order to describe a more complex concept, users may need to enter a query consisting of a mix of low and high-significance keywords. This type of query is frequently written in natural language. Conventional systems are unable to effectively identify relevant results to a natural language query. If the user does not identify a relevant result, the user may attempt to restructure the natural language query in an attempt to obtain a more relevant result set, which may further frustrate the user, as a conventional search engine is not likely to produce more relevant results in response to a restructuring of the natural language query.